Identifying People Wearing Masks in a 3D-Scene

ABSTRACT

Now people are facing the pandemic COVID-19 and have to wear masks. This brings a problem in face recognition—occlusion problem and particularly, identifying people wearing masks in 3D-scenes is a great challenge. This study aims to develop a system for tackling this challenge. The 3D-scene is constructed with the 2D-3D coordinate transformation. For the convenience of the fusion between the virtual scene and real scene, a 3D model is achieved by Sketchup Pro. The faces and masks data are explored from the video and occluded faces recognition is achieved with the convolutional neural network. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.